4 ms·

https://healthfeedback.org/claimreview/ivermectin-study-itajai-contains-methodological-weaknesses-questionable-conclusions/ https://healthfeedback.org/claimrevi

https://healthfeedback.org/claimreview/ivermectin-study-itajai-contains-methodological-weaknesses-questionable-conclusions/ https://healthfeedback.org/claimreview/ivermectin-study-itaj... Seems like this study may have some methodological issues.

srcreigh 5y ago
I read this. I am no epidemiologist but it does seem like a pretty weak critique. The core of this critique is that ivermectin may not have been used consistently among the treatment group, therefore negative results may be masked, or confounding variables may be present. It seems that ivermectin is generally known to be safe drug for use to treat other antiviral conditions. Is the critic calling that into question? I would expect more evidence for that claim. Furthermore every science experiment can have confounding variables, so yeah. Even in that case I would expect more of a curious response as 66% reduction in hospitalization at very low p-value is very significant. What could the confounding variables be? Could they be surveyed? The authors of the article did look into some confounding variables and did find that people more at risk of covid self selected to be in the ivermectin group. I would like more info about all this. I don't think it's as simple as "bad study move on".
- tptacek 5y ago
  That's one of several critiques of the study; others includes the fact that it's not controlled for confounders, that the city in question had the highest COVID mortality in its state during the program, and that several of the study authors have either conflicts, bad reputations, or both. The Twitter link itself links to even more damning critiques, including some evidence of falsified data.
  - srcreigh 5y ago
    I read the city had high mortality from the beginning of the pandemic[0]. Would it be a bad case if the population happens to be more aged etc? Is the falsified data just the off by one in deaths among the under 30 cohort or is there more?[1] [0]: https://mobile.twitter.com/sean_purdy/status/1470416523076542464 https://mobile.twitter.com/sean_purdy/status/147041652307654... [1]: https://mobile.twitter.com/GidMK/status/1478975913094090752 https://mobile.twitter.com/GidMK/status/1478975913094090752
    - snovv_crash 5y ago
      [0] could also be due to a high incidence of worms in the population.
  - torstenvl 5y ago
    > That's one of several critiques of the study; others include[] . . . that several of the study authors have either conflicts, bad reputations, or both. That is not a valid critique of the study. It is a potentially valid reason to adjust your Bayesian priors applicable before reading the study, but is utterly nobprobative as to whether the science is sound.
    - FabHK 5y ago
      Agreed. When Pfizer studies the effectiveness of their vaccine, they also have conflicts. That doesn't impugn the results of the study, if it's well done.
      - avs733 5y ago
        No, it is a critique. When Pfizer does it it is a critique as well. The question is how much of a critique. In general… If you are dishonest about your conflict, you hide it, don’t do anything to control for it then having a conflict is a bigger critique. If you are transparent about your conflict, honest about what it is, and take measures to control for it or add extra steps to analysis to prevent potential bias - then a conflict is a less strong critique of the study. Critique in scientific research isn’t a binary. It isn’t “there’s one critique and it’s valid or not and if it’s valid the study is bad”. There can be critiques of ANY study. No study is perfect. But the types and the details of the critiques matter collectively to make the study better or worse. Collectively, this study falls to the worse end. It’s not just the conflict of interest it’s the conflict of interest + the study design. The critiques of the study design are that this study is design almost exactly the way you would if you wanted to increase the appearance of effectiveness of this treatment.
  - naasking 5y ago
    > that the city in question had the highest COVID mortality in its state during the program This is not a critique of the study really. The high mortality could be due to other factors, like an older population or less careful application of other COVID mitigations. If the latter, this study would actually be a good test case for prophylactic Ivermectin actually because the signal from any effect would be stronger.
- tomlockwood 5y ago
  > Is the critic calling that into question? Are they? > The core of this critique is that ivermectin may not have been used consistently among the treatment group, therefore negative results may be masked, or confounding variables may be present. Yes. > I would like more info about all this. I don't think it's as simple as "bad study move on". Ok: https://twitter.com/GidMK/status/1425297498042556423 https://twitter.com/GidMK/status/1425297498042556423 Or was that not the info you meant?
- plouffy 5y ago
  > I am no epidemiologist Thank you for adding that, didn't have to read any further.
- FabHK 5y ago
  The main point of the critique, that people didn't take it consistently, is actually strengthening the study, isn't it? Even with only a partial treatment, you still see a significant effect. (Having said that, I don't think there's much evidence for Ivermectin overall, but this specific critique of the study seems weak.)
  - avs733 5y ago
    Not quite. My wife’s an epidemiologist (and a damn grouchy methods person…I get yelled at about study design A LOT). Her first check on any study like this is the denominator of different groups because that drives your results. This study has sketchy as hell denominators. They artificially inflate the denominator of the ivermectin group while also shifting infections to the non ivemecrim group…that skews the infection statistics by increasing the denominator artificially. The critique is basically saying “you are using the 139k people who said ‘sure’ when offered drugs as the denominator and you shouldn’t” If 75% of those people didn’t take the dang drugs (that’s a hypothetical not a claim). Then you 44% decrease in infections evaporates. This is terrible study design because of how they grouped people. They didn’t do follow up they didn’t ask whether people took the drugs (from what I can see) they didn’t have a placebo group of any form…they just offered people drugs and used those numbers to draw broad conclusions.

srcreigh 5y ago

I read this. I am no epidemiologist but it does seem like a pretty weak critique. The core of this critique is that ivermectin may not have been used consistently among the treatment group, therefore negative results may be masked, or confounding variables may be present. It seems that ivermectin is generally known to be safe drug for use to treat other antiviral conditions. Is the critic calling that into question? I would expect more evidence for that claim. Furthermore every science experiment can have confounding variables, so yeah. Even in that case I would expect more of a curious response as 66% reduction in hospitalization at very low p-value is very significant. What could the confounding variables be? Could they be surveyed? The authors of the article did look into some confounding variables and did find that people more at risk of covid self selected to be in the ivermectin group. I would like more info about all this. I don't think it's as simple as "bad study move on".

tptacek 5y ago
That's one of several critiques of the study; others includes the fact that it's not controlled for confounders, that the city in question had the highest COVID mortality in its state during the program, and that several of the study authors have either conflicts, bad reputations, or both. The Twitter link itself links to even more damning critiques, including some evidence of falsified data.
- srcreigh 5y ago
  I read the city had high mortality from the beginning of the pandemic[0]. Would it be a bad case if the population happens to be more aged etc? Is the falsified data just the off by one in deaths among the under 30 cohort or is there more?[1] [0]: https://mobile.twitter.com/sean_purdy/status/1470416523076542464 https://mobile.twitter.com/sean_purdy/status/147041652307654... [1]: https://mobile.twitter.com/GidMK/status/1478975913094090752 https://mobile.twitter.com/GidMK/status/1478975913094090752
  - snovv_crash 5y ago
    [0] could also be due to a high incidence of worms in the population.
- torstenvl 5y ago
  > That's one of several critiques of the study; others include[] . . . that several of the study authors have either conflicts, bad reputations, or both. That is not a valid critique of the study. It is a potentially valid reason to adjust your Bayesian priors applicable before reading the study, but is utterly nobprobative as to whether the science is sound.
  - FabHK 5y ago
    Agreed. When Pfizer studies the effectiveness of their vaccine, they also have conflicts. That doesn't impugn the results of the study, if it's well done.
    - avs733 5y ago
      No, it is a critique. When Pfizer does it it is a critique as well. The question is how much of a critique. In general… If you are dishonest about your conflict, you hide it, don’t do anything to control for it then having a conflict is a bigger critique. If you are transparent about your conflict, honest about what it is, and take measures to control for it or add extra steps to analysis to prevent potential bias - then a conflict is a less strong critique of the study. Critique in scientific research isn’t a binary. It isn’t “there’s one critique and it’s valid or not and if it’s valid the study is bad”. There can be critiques of ANY study. No study is perfect. But the types and the details of the critiques matter collectively to make the study better or worse. Collectively, this study falls to the worse end. It’s not just the conflict of interest it’s the conflict of interest + the study design. The critiques of the study design are that this study is design almost exactly the way you would if you wanted to increase the appearance of effectiveness of this treatment.
- naasking 5y ago
  > that the city in question had the highest COVID mortality in its state during the program This is not a critique of the study really. The high mortality could be due to other factors, like an older population or less careful application of other COVID mitigations. If the latter, this study would actually be a good test case for prophylactic Ivermectin actually because the signal from any effect would be stronger.
tomlockwood 5y ago
> Is the critic calling that into question? Are they? > The core of this critique is that ivermectin may not have been used consistently among the treatment group, therefore negative results may be masked, or confounding variables may be present. Yes. > I would like more info about all this. I don't think it's as simple as "bad study move on". Ok: https://twitter.com/GidMK/status/1425297498042556423 https://twitter.com/GidMK/status/1425297498042556423 Or was that not the info you meant?
plouffy 5y ago
> I am no epidemiologist Thank you for adding that, didn't have to read any further.
FabHK 5y ago
The main point of the critique, that people didn't take it consistently, is actually strengthening the study, isn't it? Even with only a partial treatment, you still see a significant effect. (Having said that, I don't think there's much evidence for Ivermectin overall, but this specific critique of the study seems weak.)
- avs733 5y ago
  Not quite. My wife’s an epidemiologist (and a damn grouchy methods person…I get yelled at about study design A LOT). Her first check on any study like this is the denominator of different groups because that drives your results. This study has sketchy as hell denominators. They artificially inflate the denominator of the ivermectin group while also shifting infections to the non ivemecrim group…that skews the infection statistics by increasing the denominator artificially. The critique is basically saying “you are using the 139k people who said ‘sure’ when offered drugs as the denominator and you shouldn’t” If 75% of those people didn’t take the dang drugs (that’s a hypothetical not a claim). Then you 44% decrease in infections evaporates. This is terrible study design because of how they grouped people. They didn’t do follow up they didn’t ask whether people took the drugs (from what I can see) they didn’t have a placebo group of any form…they just offered people drugs and used those numbers to draw broad conclusions.